How to Define an Ideal Customer Profile for a New Export Market

18 August 2026

Your domestic ICP does not survive the border

Most Swiss SMEs already know who their best customers are at home. Ask a founder to describe them and you'll get a confident, specific answer: this industry, this size, this kind of production problem, usually introduced through a distributor or a trade fair.

Then the same company enters Germany, or Austria, or the Netherlands, and quietly assumes that profile travels with them. It rarely does — not because the product changed, but because everything around the product did.

At home, your best customers may have found you through twenty years of word of mouth. That reputation does not cross the border. A domestic buyer might tolerate a four-week lead time because they know you'll pick up the phone; a new foreign buyer has no reason to extend that trust. Your Swiss customers may skew toward companies of 30–80 employees because that's the shape of the Swiss industrial base — but in Germany, the equivalent buying behaviour might sit at 150–400 employees, because the market's whole size distribution is different.

The profile that describes who already buys from you is a description of your history. What you need for a new market is a hypothesis about who can buy from you there — built deliberately, then tested.

Start from evidence, not ambition

Before you write anything forward-looking, look backwards. Export your last three years of customers into a spreadsheet — every one, not just the flattering ones — and add columns for: industry, employee count, what they bought, what they paid, how long the sales cycle took, how much support they consumed after the sale, and whether they're still a customer.

Then sort by gross margin per customer, not by revenue. The pattern that matters is almost never the largest logos. It's the segment where you win often, deliver without strain, and don't discount.

Two questions do most of the work:

Where do you win disproportionately? Find the segment where your close rate is well above your average. That's where your product genuinely fits, as opposed to where you have simply been persistent.
Where do you deliver without pain? Find the segment with the fewest support escalations, the fewest custom modifications, the fewest late deliveries. Profitable customers and easy customers overlap more than most teams expect.

The intersection of those two is your real strength. It is the only defensible starting point for a foreign hypothesis.

The five dimensions

A usable export ICP is specific enough that two people on your team would build the same target list from it independently. Five dimensions get you there.

1. Firmographics — but calibrated to the target market. Industry, sub-sector, employee count, revenue band, ownership structure. The trap is copying your Swiss bands directly. Look up the actual size distribution of your target industry in the target country and adjust. A "mid-sized manufacturer" means something different in Germany, Austria, and Poland.

2. Technical fit. What must be true about their operation for your product to work at all? Machine types, materials, throughput, existing systems, certification requirements, tolerances. This is the hardest filter to fake and the most valuable one — it eliminates companies that would waste a full sales cycle before discovering an incompatibility.

3. Buying trigger. What event makes a company start looking for a supplier like you? A new regulation, a failed supplier, a capacity constraint, a plant expansion, a quality incident, a tariff change that made their current source expensive. Companies with a trigger buy in months. Companies without one buy in years, if ever. If you can't name the trigger, you're not targeting — you're broadcasting.

4. Procurement maturity. How does this company actually buy? A 60-person family manufacturer where the owner decides is a completely different sale from a 400-person firm with a formal supplier-qualification process, an audit requirement, and a purchasing committee. Same industry, same size band, entirely different sales motion, timeline, and required documentation. Decide which you're equipped to handle before you build the list.

5. Serviceability — the export-specific dimension. This is the one domestic ICPs never contain, because at home the answer is always yes. Can you actually serve this customer profitably from Switzerland? Consider shipping cost relative to order value, customs and documentation burden, whether your certifications are recognised in that market, whether you can meet expected response times across a distance, which language your support and documentation exist in, and whether your warranty terms are workable.

A customer you can win but cannot serve profitably is worse than no customer. Serviceability is what turns an attractive market on paper into a sustainable one, and it deserves equal weight with the other four.

Turn the profile into a searchable definition

An ICP written in prose is a discussion document. An ICP written as filters is a target list. Translate each dimension into something you can actually query:

Industry becomes NACE codes (Europe) or SIC codes, not adjectives.
Size becomes employee and revenue ranges with explicit floors and ceilings.
Geography becomes regions, not countries — industrial demand clusters geographically, and a national list wastes most of your effort. In Germany, machinery and automotive concentrate in the south and west; in Austria, industry concentrates in Upper Austria, Styria and Vorarlberg rather than Vienna.
Technical fit becomes observable signals: equipment mentioned on their site, certifications listed, materials named in their product pages, job postings for specific skills.
Trigger becomes a monitorable event: expansion announcements, new-plant news, regulatory deadlines, leadership changes in procurement or operations.

Also write the exclusions explicitly. Companies too small to afford you, too large to care, in adjacent sub-sectors that look similar but buy differently, or in regions you can't serve. A negative list prevents more wasted work than a positive one, and it's the part teams most often skip.

Finally, map the buying roles. In industrial B2B the person who first evaluates you is rarely the person who signs. Name both — typically a technical evaluator (production, engineering, quality) and a commercial approver (purchasing, managing director) — because they need different messages, and a sequence written for one will fall flat with the other.

Test the hypothesis before you scale it

Your first export ICP is a guess. Treat it that way.

Build a deliberately small list — 50 to 100 companies that match tightly — and run real outreach against it. Then read the results as diagnostics rather than as a verdict on the market:

Low reply rate across the board usually means the message or the trigger is wrong, not the profile.
Replies that say "interesting, but not for us" mean your technical-fit filter is too loose. Tighten it.
Good conversations that stall before purchase usually mean you targeted the wrong role, or misread procurement maturity.
Meetings that convert — study these hardest. Two or three won deals in a new market tell you more about the real profile than any amount of desk research.

Revise the profile every time you learn something, and keep the revisions written down. After one quarter of disciplined outreach you'll have an ICP grounded in that market's evidence rather than your home market's history.

The mistakes that cost the most

Targeting a country instead of a segment. "We're expanding to Germany" is not a plan. Germany contains thousands of segments and you can serve a handful. Choosing the market and choosing the profile are two different decisions — if you haven't done the first one systematically, our guide to choosing your next export market covers the scoring framework.

Writing the ICP so broadly it excludes nobody. If your profile admits 40,000 companies, it isn't a profile. Narrow until it feels uncomfortably specific, then run it. You can always widen later; you cannot recover the months spent on a list that was never qualified.

Skipping serviceability. The most common failure in Swiss export sales is winning customers that cost more to serve than they generate, discovered only after the third shipment.

Never revisiting it. An ICP written once and filed is worth almost nothing. The version that matters is the one on its fourth revision, shaped by real replies.

From profile to pipeline

Once the profile is specific and testable, everything downstream gets easier and cheaper. You know which companies to source, which contacts to verify, which trigger to reference in a first email, and which signals mean a company is worth a follow-up.

That's the practical payoff: a tight ICP doesn't just improve targeting, it makes every subsequent step — buyer discovery, contact verification, sequence writing, follow-up prioritisation — measurably more efficient. If your next step is building the list itself, our guide on finding B2B buyers in Germany walks through the channels, and the cold email guide for exporters covers turning that list into conversations.